genesis/api_clients/pubmed_api.py

# genesis/api_clients/pubmed_api.py
import os
import requests
from typing import List, Dict, Optional
from xml.etree import ElementTree as ET

NCBI_API_KEY = os.getenv("NCBI_API_KEY")  # Optional but increases rate limits
NCBI_BASE = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils"

def search_pubmed(query: str, max_results: int = 10) -> List[str]:
    """
    Search PubMed and return a list of PMIDs.
    """
    params = {
        "db": "pubmed",
        "term": query,
        "retmax": max_results,
        "api_key": NCBI_API_KEY
    }
    r = requests.get(f"{NCBI_BASE}/esearch.fcgi", params=params)
    r.raise_for_status()
    root = ET.fromstring(r.text)
    return [id_tag.text for id_tag in root.findall(".//Id")]

def fetch_pubmed_details(pmids: List[str]) -> List[Dict]:
    """
    Fetch detailed information for a list of PMIDs.
    """
    if not pmids:
        return []
    
    params = {
        "db": "pubmed",
        "id": ",".join(pmids),
        "retmode": "xml",
        "api_key": NCBI_API_KEY
    }
    r = requests.get(f"{NCBI_BASE}/efetch.fcgi", params=params)
    r.raise_for_status()
    
    root = ET.fromstring(r.text)
    articles = []

    for article in root.findall(".//PubmedArticle"):
        title = article.findtext(".//ArticleTitle", default="No title")
        abstract = " ".join([t.text for t in article.findall(".//AbstractText") if t.text])
        journal = article.findtext(".//Title", default="Unknown Journal")
        pub_date = article.findtext(".//PubDate/Year", default="Unknown Year")
        doi = None
        for id_tag in article.findall(".//ArticleId"):
            if id_tag.attrib.get("IdType") == "doi":
                doi = id_tag.text
        authors = []
        for author in article.findall(".//Author"):
            last = author.findtext("LastName")
            fore = author.findtext("ForeName")
            if last and fore:
                authors.append(f"{fore} {last}")
        pmid = article.findtext(".//PMID")
        articles.append({
            "pmid": pmid,
            "title": title,
            "abstract": abstract,
            "journal": journal,
            "pub_date": pub_date,
            "doi": doi,
            "authors": authors,
            "url": f"https://pubmed.ncbi.nlm.nih.gov/{pmid}/"
        })
    return articles

def search_and_fetch(query: str, max_results: int = 10) -> List[Dict]:
    """
    Convenience function: Search and fetch results in one step.
    """
    pmids = search_pubmed(query, max_results)
    return fetch_pubmed_details(pmids)